1. Field of the Invention
The present invention relates to an image processing device, an image processing method and a recording medium.
2. Related Art
In digital cameras, portable telephones having an image capture function, and the like, the light incident from a lens is converted to electrical signals by way of an imaging elements CMOS (Complementary Metal Oxide Semiconductor) or CCD (Charge Coupled Device) type, and these electrical signals are outputted as image data.
The aforementioned imaging elements photoelectrically convert incident light and accumulate charge, and have a plurality of pixels that determine the brightness based on the amount of accumulated charge. In this plurality of pixels, there is a possibility for white defects to occur in which a charge is accumulated that exceeds the amount according to the incident light.
Japanese Unexamined Patent Application, Publication No. 2000-101925 discloses a method of specifying in advance a pixel in CCD adopting imaging elements at which a white defect is occurring, and correcting a pixel value (image signal) corresponding to this pixel based on the pixel values (image signal) corresponding to before and after pixels.
However, in CMOS-type imaging elements, there is a possibility for blooming to occur in which the white defect expands to pixels arranged in a peripheral region to the pixel at which the white defect occurs.
FIGS. 6A to F are diagrams illustrating blooming. In FIGS. 6A to F, one square box indicates one pixel. The color of each pixel indicates the amount of charge, meaning that a pixel approaches white as charge is being accumulated.
First, in FIG. 6A, a white defect occurs in one pixel (pixel in the center of the figure). At the initial occurrence, the amount of charge accumulated more than necessary is small, and the degree of the white defect is low. However, due to environmental changes such as temperature and humidity or ageing, when the amount of charge accumulated more than necessary increases and reaches the permitted value (FIG. 6B), the charge will leak out to adjacent pixels, and a charge will accumulated more than necessary also in the adjacent pixels (FIG. 6C). Thereafter, the charge being leaked to adjacent pixels from the pixels at which a charge more than necessary has accumulated in this way occurs like a chain reaction, and the white defect grows (FIGS. 6D to F).
In order to correct the white defect occurring due to blooming, for example, a digital camera or the like repeats an image capturing action two times consecutively. In other words, the digital camera or the like captures a normal subject in a first image capturing action, and captures in a darkened state in the second image capturing action. The digital camera or the like specifies the pixel at which a charge has accumulated in the results of such a second image capturing action as a pixel in which a white defect has occurred. Then, in the image data captured in the first image capturing action, the digital camera or the like performs correction of the pixel value (image signal) of the pixel specified in this way. However, the method disclosed in Japanese Unexamined Patent Application, Publication No. 2000-101925 corrects the pixel value of the pixel at which the white defect is occurring based on the pixel value before and after; therefore, there has been concern over the correction result being unnatural.
In addition, there has been concern over the overall photography time taking too long, if a second photography action is made to specify the pixel at which the white defect occurs during photography in order to raise the accuracy in specifying the pixel in which a white defect occurs.
As a result, an image processing device and image processing method have been desired that can correct accurately and effectively correct pixel values in image data (image signals).